The Standard I Can't Meet Either

I made an episode about AI and trust. I didn't mention I'm inside the question.

I recorded an episode this week about how audiences penalize AI-generated content. See below.

I use AI tools. I didn't say that in the episode. So there’s that.

If you create content using AI tools and care deeply about human authenticity, you’re probably sitting with an inner tension that’s embarrassing to name. I’m surprised more people haven’t written about it.

Creators like me have access to generative AI tools that make creating content faster than ever. And it all looks so polished. So crisp.

And yet, when I was researching the latest findings on how audiences are receiving the massive content deluge happening right now, I found myself conflicted.

I’m a creator. I’m also an audience member. Yep, the irony wasn’t lost on me.

Research is consistently showing that audiences devalue content when they know even a tiny bit of it was AI generated. The devaluation isn’t proportional. It’s not linear.

This concept has been coined by moral psychologists as the “Contamination Problem.”

Whether a creator used AI for 1% of a piece or 40%, the audience’s reaction was the same. You can listen to the recording above for the full details on how and why.

Most creators don’t let on how much AI they use. It’s understandable. Everyone hates AI slop. There’s enormous pressure right now to keep things real, keep the human element.

But not everything made with AI is slop.

These tools let me create things I couldn’t have created before. After all, it wasn’t long ago that I laughed at AI’s attempts at writing because it was so obvious. Flat. Everyone could tell. Now the tools have improved enough that I can instruct my AI to write in my voice, my style, my rhythm.

And yet the research shows that even as the tools got better, audience reaction to AI involvement didn’t change at all.

Which means the issue was never about the quality of the output. It’s about something else entirely.

Now there’s pressure running in both directions. Audiences want authenticity. They also expect a level of polish that only AI produces at scale. Creators who don’t use AI may actually be at a disadvantage now.

It’s a bit crazy. I’ve felt this myself. For example, I’m embarrassed that my podcast doesn’t have a beautiful intro, AI-generated timestamps, perfect music interludes, etc. It just hasn’t been a priority for me to do all of that yet. I’m more interested in the content and the message.

For what it’s worth, I use AI mainly to edit what I write. I don’t use it to write my content because I don’t like the way it writes for me. Even when it’s trying to sound like me.

(OK, full confession — I did use AI to come up with the title of this article. I figured I wasn’t going through all of this to have no one read it. That’s just creator pragmatism leaking through.)

My podcast episode doesn’t resolve the tension with any great answers. I was just describing what I notice is lurking beneath things. And if you got this far, you’re probably someone who feels it too.

Thanks for listening. .